Hey all:

I am looking for a second opinion on this attached report...my customer is more worried about it than I anticipated.

I sent out the images in this report, along with a reccomendation to inspect the joint for corrosion and torque at the next shutdown, in June. Thermographs will be taken periodically from now until then.
This is a main bus that feeds one of our buildings. One of the joints shows about a 5c rise. Now normally we don't report 5c rises, but I did in this case because I am uncertain as to the structure underneath the duct work, and this building is critical.

Thermographic_Report_for_B3B_Main_Bus.doc (805 Kb, 77 downloads)

Martin,

Are we looking at exposed bus, or at the ductwork surrounding the bus? Your report states, "The 2 dots at the center of the rectangle are the bolts clamping the bus together. Being in direct contact with the bus, these bolts are also a good indicator of the true condition of the bus." I'm not familiar with this installation, but most bus ducts are enclosed and grounded, with periodic insulated supports to maintain the energized bus centered in the duct. The electrical insulation is typically also a thermal insulator, and often masks the true temperature of the hotspot on the energized bus, when viewing the outside surface of the duct. Speaking generally, even minor temperature rises on enclosed bus duct are treated as potentially serious hotspots.

It is important to understand the mechanism of heat transfer between the point source of heat (the energized bus bolted connection) and the observable emitting surface (painted ductwork).

Also, did you verify the emissivity of the ductwork, and take into account the reflected background? When you are outdoors and look up at an angle on a non-blackbody, your reflected background is the sky. In a clear sky, that is a very low reflected contribution of energy; in a cloudy sky, it is significantly higher.

If you actually have a line-of-sight onto the bus itself, know that the emissivity of bus is typically very low, and that the actual temperature is often quite higher because of the reflected component.

I would be concerned about this hotspot, and would carefully evaluate the heat transfer characteristics of the system before dismissing it as a minor issue.

Rich Wurzbach
ASNT PdM Level III-IR and Thermal Testing
Maintenance Reliability Group, LLC

The other basis of my opinion here is the narrow width of the duct work. Including both busses, insulation and Ductwork, it is less than 2 " thick. I have not performed an emmisivity test, but the surface looks to me like it has fairly high emmisivity. There are corroded sections on the duct work and support structures, which we know to have high emmisivity, which do not stand out in the thermographs. Based on that I would be suprized if intact duct surface has a significantly lower emmisivity.

One of my concerns with this anamoly is the distance from the anamoly that I had to go to get a reference point on the same phase.
Although the delta is only 5C, the energy radiates quite a distance... 20' 30'??

We did not cover this idea in Thermography Level 1, so I would like some opinoins on the significance of the distance heat radiates from an anamoly.

Obviously, more energy is required to heat up more bus.

To answer your question, this is ductwork around the bus question.

"One of the things you said caught my attention:
The electrical insulation is typically also a thermal insulator, and often masks the true temperature of the hotspot on the energized bus, when viewing the outside surface of the duct."

The bolts go entirely through both sides of the bus-bar connection. Any anamoly would presumably be at the connection. So I would think that if we had and significant anamoly at the melting point of copper (1083.0 °C ) that the bolt in close proximity to it would be greater than 20 C.

I suppose if it were a very small surface area inside the connection it could have this effect.

1 more thing: we are 1327 amps on a 4000 amp Bus.
Stable load, no expectation of a much higher load

I wrote a response but it seems to have disappeared.

Just a remote possibility - sometimes outdoor bus ducts have thermostatically controlled space heaters which switch on in cold weather, even when bus is energized. I call it remote because I don't see any external conduit entering here that would be a tipoff that there's a heater inside.

With that high current, inductive heating can always be considered, but I don't really see anything to suggest that.

If it is a bus high resistance connection, than it is indirect heating and there is more reason to be concerned because it's very difficult to estimate reliably the temperature inside. Regardless of whether the bolts do direct to conductor it is a concern but more so if there is an insulator. If there is no electrical insulator involved then those external bolt heads would have to be energized. That doesn't seem reasonable to me but you never know. What voltage is the system by the way.

Is it possible to get a shot from the back or is the duct too close to the building?

The pictures look like there is a ridge sticking out at the hot spot. Just like the corners in a room some ridge areas are hotter or colder because of air flow problems.

Yes, there is a small ridge there. No cables for a heater. The duct is less than 6" away from the building, but I could get a reflector and angle in to look at the back.
I used The CSI Sonic scan model 7000 this morning, and could not detect any arcing. But the range was 15-20', so I'm not sure if arcing could be detected at that distance

Also:
I would like the group to address this question:
Independent of the max temp of the anamoly, is the distance between the max temp rise and the point it again is at nominal for load indicate severity?
ie, the distace the heat from the anamolie conducts from the source should be an indicator of severity ... it takes more energy to conduct heat a further distance. Based on this, the anamolie would be classified as severe, since clearly the heat rise is many feet long.

Of course the problem is to determine what is "nominal". Here I am defining "nominal" as the temperature on other sections of the same phase of the same duct (same load and same conductor) distant enough from the anamoly to exclude any heat generated there.

quote:

would like the group to address this question:
Independent of the max temp of the anamoly, is the distance between the max temp rise and the point it again is at nominal for load indicate severity?
ie, the distace the heat from the anamolie conducts from the source should be an indicator of severity ... it takes more energy to conduct heat a further distance. Based on this, the anamolie would be classified as severe, since clearly the heat rise is many feet long.

I think the answer is yes. Most connections we look at are very close to the location of the high resistance connection and we can assume the connection itself is not much hotter. When the connection is enclosed somehow (inside oil filled breaker, transformer LTC, bus duct etc) and we are looking at the outside of the enclosure, there is a lot of thermal insulation which will create a temperature difference (gradient) between the contact and the external surface. Most people call this indirect heating. 5C viewed on external surface with indirect heating would be much more SEVERE than 5C if you are looking at the metal directly associated with the connection.

Snell has published some good articles on prioritizing electrical anomalies. Here is one:
http://snellinfrared.com/tt/news0703/July-2003_TT.pdf

Read the article titled "So many problems, which one to fix?". There is a "stage 1 question" "Are electrical gradients large to moderate". The associated note reads "Such as a bus stab, enclosed bus connection, motor terminal box, massive
connection, or connections inside any oil-filled device.". In Snell's ranking system, an answer "yes" to this question (indicating indirect heating) kicks you into a higher severity ranking.
=========
(edited to add "severe" that I left out before)

This message has been edited. Last edited by: electricpete,

I did an emmisivity test on the duct work, and in fact standard electrical tape cannot even be seen in the thermographs, so the temperature reading on the ductwork is fairly accurate

Martin,

I agree with e-pete's comments. With regard to induction heating, in symetrical bus duct, this would typically only happen if the bus was not centered in the duct, which in itself is a significant problem. I continue to believe this issue requires high priority until further investigation reveals otherwise.

Rich

Just a little more on the subject of "indirect heating"

I was looking for this file yesterday and couldn't find it:
http://www.flirthermography.com/media/029snell.pdf

It is another article by Snell on the same subject but with a lot more details. It starts out talking about the available standards that are based strictly on temperature rise. Then he points out the many flaws of relying only on temperature rise. But rather than giving up and saying "Only an expert can decide", he gives a more detailed set of rules that someone might use to evaluate a condition considering many other factors besides just temperature rise. One of the factors in his matrix is labeled "Is thermal gradient large?". High thermal gradient is his terminology for what most people call "indirect heating".

I really like this type of approach. Sure an expert opinion is considered by most thermographers to be the best, but folks at our plant won't settle for something so subjective as an opinion. They want a somewhat objective set of rules to help us consistently communicate the severity of our findings and prioritize repairs. I'm hoping to get together with our plant stakeholders and use these articles as a starting point develop a similar approach for our plant.

Thanks Pete.. you rock!

Pete:
I read the paper, and on pg 2 fig 3 he refers to the Thermal gradient as the delta between the 2 ends of the splice, not the gradient between max and the nominal temp of the conductor at a given temperature.
So I think Thermal gradient is a general term used to dercribe the difference between two points in a given thermal system...

Anyway, in my case not knowing what the true temperature is inside the bus, the gradient is unknown. The gradient I was referring to was 2 points, chosen on the ducting in the thermographs. The reference point is a point on the same phase of the anamoly, at a point where the temperature appears stable and not influenced by the anamoly. The other point is the maximum temperure as a result of the indirect heating on the ducting.
My question was this:
A certain amount of watts are required to heat up a given mass. The more mass, the more heated required to reach a certain temperature. In addition, the more heat, the further away from the source of the heat the mass is conducted.

That is the thing which gives me pause about this anamoly. The distance the heat conducts along the duct, before it reaches a stable "nominal" temperature.
In fact, I am more concerned about this than I am the indirect heating issue, because the entire width of the assembly, bus, insulation and ducting, is only 1.75 inches.

Woops: The question was this. Is this distance described above indicitive of severity?
I think it is.

MMM

Thanks for the compliment (I rock). I wish my kids would think that way. It's worthwhile to mention that Rich was the one who immeidatley recognized this as indirect heating.

Sorry that I misrepresented the contents of Snell's paper. You are right that he is not using the term thermal gradient in exactly this way. I'm not sure exactly how Snell's system will boost the priority for indirect heating scenarios (maybe the unreliable temperature estimate category?), but clearly it IS appropriate that any case of indirect heating (where we are not looking directly at the contact or connecteed metal) should be viewed much more severely. Trying to apply the standard temperature rise rules to the external surface temperature of an enclosed connection is a recipe for disaster. I agree wholeheartedly with Rich on that.

I think that the presence of indirect heating (insulation between a presumed hot connection on the bus) and the visible outside of the enclosure is the main issue. The only "distance" that should concern us imo is the distance associate with that thermal insulation.

I understood your most recent question associated with the following aspect. The external surface temperature is highest at one point, and then "tails off" (slowly decreases) as we move up or down away from that point on the bus enclosure surface. I think maybe you are concerned that it does not tail off fast enough?... i.e. we still see evidence of the elevated temperature a few feet above and below the hot spot.

I don't really think that aspect is telling us anything about the severity for two reasons:
1 - You have a relatively small temperature span (appropriate in this case) which somewhat exagerates how far the effects of the heating at the center spread up and down.
2 - Tailing off behavior is more a function of heat transfer characteristics than fault severity. Let me explain why I think this. My model would be that heat from the bus is conducted through the porcelain insulator to the bus enclosure at the center of your temperature distribution. I assume the enclosure itself is thin and high thermal condcutivity so it doesn't have any gradient from inside surface of enclosure to outside surface of enclosure. Once the heat reaches the enclosure it flows two ways: 1 - up and down the enclosure due to thermal conduction and 2 - perpendicular away from the enclosure cover into the atmosphere by convection and radiation. These two simultaneous effects create a tailing off. Which effect is stronger determines how fast it tails off. If the conduction up and down the enclosure is high and the convection/radiation to atmosphse is low, then the tailing off is very gradual. If the conduction up and down the enclosure is low and the convection/radiation to environment is high then you see a rapid tailing off and the hot spot doesn't "spread out" very far. Tailing off away from a hot-spot in general follows somewhat an exponential characteristic: T(x) =Tinfinity + (Tcenter-Tinfinity)*exp(-Lamba*x) where x is distance (height in this case), Tcenter is the hottest spot, Tinfinity is the temperature of the coolest spot and lambda is a paramteter which depends only on the heat transfer characteristics of the system. Lambda is directly proportional to linearized film coefficnet and radiation coefficient which carry heat away from the surface to the ambient. Lambda is inversely proportional to thermal conductivity of the metal duct enclosure along it's lenght. The only place that the fault severity comes into that temperature distribution equation is heat generation which drives Tcenter - Tinfinity. The way that the temperature distribution tails off is dependent on the Lambda parameter which depends only on heat transfer parameters and is generally not an indication of severity imo. I can provide more detailed math proof of the exp(lambda*x) discussion if you'd like.

All of my above discussion based on the assumption that we are in fact generating heat at a hot connection of the bus which is conducted through insulator to enclosure at that one point. The temperature distribution and tailing off are still a very important apsect to the extent that it helps us understand whehter that assumption is correct or there might be some other cause. Looking again at the very first picture, it does look like the center phase enclosure temperature stays relatively steady for a distance above the hot spot. And then gets hotter at the next insulator station above. Can you draw a vertical temperature profile?

Maybe this is telling us that heat is being conducted along the bus to the upper insulator station as well. That would not be surprising since the bus is a great heat conductor and therefore would not change the diagnosis. But I may be off-base - best to keep an open mind to try to understand what the picture is telling us.

I'd like to wind up my rambling by repeating that I agree with Rich on the severity.

This message has been edited. Last edited by: electricpete,

On the subject of interpretting Snell's terminology, in the first linked item (Thinking Thermally newsletter), he clearly used "thermal gradient" to refer to what we have been calling indirect heating. As quoted above he gives examples "Such as a bus stab, enclosed bus connection, motor terminal box, massive
connection, or connections inside any oil-filled device." The enclosed bus connection, motor terminal box and stuff inside oil-filled device are all clear-cut examples of what we have been calling "indirect heating." The bus stab presumaly is at a difficult point to inspect directly so you have to inspect bus nearby? Massive connection presumably may have the connection itself far down inside the surface (although if it's all solid metal it should conduct the heat fairly well).

Bottom line is Snell uses the term thermal gradient to mean indirect heating in his earlier article and specifically calls out the example of enclosed bus.

Pete:
"My model would be that heat from the bus is conducted through the porcelain insulator to the bus enclosure at the center of your temperature distribution"

In your experience do most buses have porcelain insulators?
This 4000 amp.

OK, I just got off the phone with Seimens, and there is no ceramic in the bus ducting. It is multilayer duct, 3 Phases and a Neutral, with 2 layers of PVC insulation between each.

So that would confirm that there is certainly a potential that this anamoly is several layers deep inside the connection, and may be considerably hotter than the ducting. The folks at Siemens said they would expect it to be from 10-20C hotter.